Smart Farming Based Early Classification of Paddy Blast Disease Using Adaptive Deep Learning Algorithm

The backbone of the Indian economy, agriculture is essential to Indians' standard way of life. Plant disease monitoring by hand typically involves a tremendous labor and takes more time to perform. The primary goal of this research is to help farmers, particularly those who are dealing with blast disease in rice crops, by quickly and reliably detecting plant diseases. The primary goal of the research is to catalog the diseases that affect rice and identify crop infections early on. Therefore, the development of soft computing and image processing technologies for smart agriculture applications is the subject of this research study, with a special emphasis on applications in innovative design for the detection of diseases in rice. The key methods of analysis used include preprocessing, classification, feature extraction, and picture acquisition. The lands used for agriculture are the source of input images. Dmey wavelet filtering is the first step in the pre-processing procedure which removes noise from images. Next, image-based crop feature extraction techniques are utilized to perform an extensive study. The proposal is to use the Adaptive deep neural network (ADNN) to address the overlapping issue. In ADNN, the total accuracy is 97.3%. A wide range of rice blast diseases are correctly identified with the desired accuracy, regardless of the obscured settings, according to the evaluation results of the suggested method.